February 21, 2011

It does pay to go to a more exclusive college, after all

One of the better known social science studies of recent years is the 2002 study by Stacy Dale and Alan B. Krueger finding that it doesn't matter what college you attend in terms of how much money you'll make after you adjust for SAT scores and high school GPA.

Actually, as Half Sigma likes to point out, that isn't what Dale and Krueger found in their study of 1976 students, that's only true when they adjusted for SAT scores and GPA and what I would call ambitiousness (by taking into account colleges applied to).  

Now Dale and Krueger are back, looking at the 2007 incomes of people who applied to a couple of dozen selective colleges in 1989:
Specifically, for the 1976 and 1989 cohorts, attending a college with a 100-point higher SAT score [on a 400 to 1600 scale] lead [sic] to students receive about 6 percent higher earnings (in 1995 and 2007 respectively) according to results from the basic model ...

On the other hand:
Indeed, the finding that the average SAT score of the highest ranked school that rejected a student is a much stronger predictor of that student’s subsequent earnings than the average SAT score of the school the student actually attended should give pause to those who interpret conventional regression-based estimates of the effect of college characteristics as causal effects of the colleges themselves. ...

So, applying to Harvard is a better predictor of future income, all else being equal, than attending Harvard. I suspect that how much money somebody makes correlates to a surprising extent with how much money they expect to make and/or think they deserve to make, which, in turn, does correlate with applying to Harvard.
About 35 percent of the students in each cohort in our sample did not attend the most selective school to which they were admitted.

21 comments:

Anonymous said...

Did the study mention the slected groups choice of college/prestige

Simon in London said...

I'd be interested in an estimate of true costs of attending eg Harvard, compared to typical future earnings boost. Going to University at all seems potentially a bad idea, with 4+ years of no or minimal earnings, even before factoring in the fees, which can be enormous.

Anonymous said...

Then, ultimately, is it the ambitiousness, work ethic, and IQ that drive who succeeds? That's what the results would seem to say.

Anonymous said...

http://calnewport.com/blog/2008/08/15/does-where-you-go-to-school-matter-and-why-reporters-get-this-wrong/

As the Half-Sigma blog puts it, the Dale and Krueger report is “possibly the most mis-cited study ever.”

I agree.

I recently re-read the report, and I came away with the exact opposite conclusion as Gregg Easterbrook. In my opinion, Easterbrook clearly cherry-picked the results to support the answer he wanted to find.

The statistic he cites has to do with students who got accepted both to schools with high average SAT scores and those with low average SAT scores, and then chose to attend schools with the low scores. As the researchers note, however: “the average SAT score is a crude measure of the quality of one’s peer group.” By contrast, when they analyzed the selectivity of the college — i.e., position in the Barron’s College Guide rankings — Dale and Kruger found:

"Men who attended the most competitive colleges earn 23 percent more than men who attended very competitive colleges, other variables in the equation being equal."

Remember, this is a regression analysis. This means that the researchers are looking at students who were accepted at both highly ranked and less highly ranked schools, and then found that those who chose to attend the higher ranked schools earned significantly more than those who chose the lower ranked schools. The researchers argue that ranking is a better indicator of selectivity than average SAT score as it better captures society’s common understanding of the school’s prestige.



Dale and Krueger originally found that, controlling for other variables, attending a highly selective institution resulted in much higher income. Attending a high SAT school did not. So even the original study suggested that, yes, you should attend Harvard.

Sylvia said...

Location is key. If you attend a top NYC college (like NYU), you're usually put on the advanced pipeline for jobs in the city (within industries such as media, don't know about others). What usually happens is that a huge number of the girls responsible for hiring in the city are NYU/Columbia grads, and they favor their own over out-of-towners. Because you go to the same bars and clubs, you also get to know the women at the firms who didn't attend college in the city. You then in effect start working for a company unofficially while still a student. They actually give you a (very limited) expence account and can help you out with student debt. They basically take you under their wing, developing you into what the company wants. It's all about networking like crazy. They also help you out socially, telling you which guys to avoid (known players), using the company name to get you on guestlists etc. All in all, for a female who wants to work in NYC, it's better to graduate in the city than at a New England ivy. I don't know if it's the same for guys.

Rain And said...

"It does pay to go to a more exclusive college, after all "

The study shows the exact opposite of what your title says it does. It pays just as much to APPLY to Harvard as it does to GO to Harvard. Prestigious Universities add no advantages onto what the students bring in with them.

jack strocchi said...

The NYT reports with a straight-face:

Japan is by far one of the cheapest markets in the world,“ said Charles de Vaulx of International Value Advisers, a New York-based investment firm. “It’s so universally hated, yet it might be one of the world’s best-performing markets over the next five years.”

“So many Japanese companies are well managed from an industrial standpoint,“ he said.

Obviously the fact that finance capital investors despise an economy where the industrial companies are well-managed has somehow escaped the NYT & WSJ irony detectors.

stari_momak said...

OT, but yet more evidence of the clash of the theoretical value of 'school choice' with the actual value of having a house in a good neighborhood with good schools

"Taft High parents object to district's offer to share space with charter"

http://www.latimes.com/news/local/la-me-taft-protest-20110222,0,7487186.story

Bonus-- crooked Russian immigrants.

Now you can have ice cream in heaven said...

Did the authors use the mean alone to compare cohorts?

If so, the results would skewed by the growing winner take all fields that the elite schools feed into like finance and elite startups. It'd be interesting to see a metric on this and how it's grown and been concentrated by the HYP+ set over the past 30yrs.

While it's easy to factor out easily measurable stats like family income, I'm doubtful any study can accurately capture and factor out the effects of family wealth and networks in these comparisons. For the supra elites attending Harvard, family wealth is probably usually more deterministic than taxable family income.

HPY+ are getting the scions of the global elite that will mostly and inevitably inherit the wealth, power and networks of their families. That doesn't mean Simple Jack from Iowa is going ultimately amass as much wealth as those scions even if he is much brighter, motivated and charismatic.

I wounder how much a HYP+ degree really helps the Horatio Algers who socially don't fit into with the already anointed ones. My guess is that hardworking and talented non-elites overestimate how much a HYP+ degree will help them on average (unless they are going into something that is exclusive but not so well paid like academics).

Finally, some schools seem more highly ranked in prestige than their academic chops would seem to merit. Places like Notre Dame have excellent professional networks and school loyalty compared to arguably better academic public schools like UC Berkeley. I wouldn't be surprised grads from places like ND and Duke did better than grads from more highly ranked schools.

Half Sigma said...

I agree with what the Anonymous commenter above said, in bold type.

Here's the link to my blog post: Attending an Ivy League school results in higher income.

I now this rubs the Ayn Rand groupies the wrong way, that attending an elite college is so important to one's earning potential regardless of one's innate ability.

Anonymous said...

Half Sigma has an interesting theory, which I agree with, on IQ and income. High IQ is useful to the extent that it allows you to attend a selective college and get on a lucrative career track.... but once you've graduated from school and broken into your desired career, IQ is not an important variable for predicting success. It seems that that other factors are more relevant. Essentially life success is about educational and career tracks, with IQ broadening the number of tracks you can jump on.

I think Half Sigma even posted that earnings peak with a SAT score of 1100. He also analyzed GSS data and found that when controlled for education, IQ does not predict income.

Half Sigma, care to comment any more on this?

Anonymous said...

I should mention that the above bolded comment wasn't mine - I was quoting another blog. The below bolded comment is from HS's blog.

Charles Murray’s points, which are entirely supported by the GSS data, are that (1) IQ by itself is correlated with higher income; (2) IQ is correlated with higher educational attainment; (3) higher educational attainment is correlated with higher income. Clearly the means by which IQ correlates with higher income is through educational credentials; smarter people are more likely to get a college degree or a graduate degree. This is what the GSS data shows. But it’s the credential that employers seem to value and not the employee’s intelligence.

Anonymous said...

If there is someone here that likes to analyze GSS data, maybe you could see if there is any correlation between work ethic and income (once controlled for education)? Perhaps "hours worked per week" or some other variable that correlates with work ethic could be used. I think hard work matters quite a bit, but I'd like to see it quantified.

Jeff said...

It's a sh*tty situation for all.

Prestigious schools know how valuable they are and will limit that value through tuition (although some of the very, very top schools are more reasonable now).

Some fields look more indifferently at a prestigious education (applied science, medicine, accounting). IMHO, if your kid seeks a job in that type of field, your better off attending a less expensive school.

Half Sigma said...

Anonymous: "I think Half Sigma even posted that earnings peak with a SAT score of 1100. He also analyzed GSS data and found that when controlled for education, IQ does not predict income.

Half Sigma, care to comment any more on this?"

My findings from the GSS and other sources, such as the mis-cited Dale Krueger study are that:

(1) For people who don't have college degrees, IQ has a correlation with income and career success.

(2) For people who do have at least a college degree, then IQ has, if anything, a NEGATIVE correlation with financial success. Having a college degree, a graduate degree, being married, being an athlete in college, all are correlated with financial sucess and IQ is not. Someone with an SAT score of 1100 who somehow gets admitted to Harvard would be predicted to have better career succes sthan someone with an SAT of 1400 who gets admitted to Harvard, as hard as that is for the Ayn Randers to believe.

(3) Now point #2 doesn't take into account extreme outliers such as Bill Gates or the guys who founded Google. I think it's possible that a high IQ increases one's chances of extreme success such as becoming a centa-millionaire, but these people are so rare that they don't show up in any studies. The vast overwhelming majority of high IQ people do not have this type of success.e

Anonymous said...

I think the super high IQ types that get rich are often from software entrepreneur backgrounds. The software industry has much lower barriers to entry than other industries. Being a good software entrapranuer is also highly g-loaded, I would think, since you might need to do a lot of the programming yourself.

Other industries have higher barriers to entry. To rise, you need to climb the corporate ladder and make connections. Then you assume a leadership position or go out on your own. That more requires hard work, social skills, leadership abilities, and pedigree than a high IQ.

Anonymous said...

I think an 1100 SAT kid that gets into Harvard must be really willing to push himself hard or extra ambitious, compared to the 1400 SAT kid.

none of the above said...

Anonymous:

I'd assume the 1100 SAT kid at Harvard has a parent with a building named after him on campus. Which is indeed strongly correlated with financial success later in life....

Half Sigma said...

A person who is just smart enough to get into and graduate from medical school is going to, on average, going to have much higher income than someone a lot smarter who doesn't go to medical school.

Sure, investment bankers and BIGLAW partners make more money than doctors, but only a tiny percentage of people smart enough to get into and graduate from medical school become BIGLAW partners or investment bankers.

M said...

What college you attended may not effect how much money you make, but may effect how you make it.

But to me, I think its pretty clear that elite universities don't add much value in terms of skill to their graduates (or else they would be compensated beyond their ability).

The legitimate purpose and selling point of elite universities is research (humanities and STEM) - elite universities admission should be oriented around determining which high school students perform the best original research, to some extent irrespective of their g or SAT scores, which although they are predictive of many things are not clearly indicative of original research ability, and then fast tracking them into research programs.

I suspect that how much money somebody makes correlates to a surprising extent with how much money they expect to make and/or think they deserve to make, which, in turn, does correlate with applying to Harvard.

Yes, and how good they are at persuading people that they deserve to make that much money, which again correlates well with making applications to elite universities (attempting to persuade an elite university to admit you probably tracks overall confidence in your own persuasion skills).

Jeff said...

In the future we will likely have to pay doctors less, which means we'll have to lower the barriers to become a doctor (cheaper and shorter medical schools...perhaps try to make it a 2 or 3 year degree).